Evaluation of postsurgical behavioral changes in patients with Parkinson’s disease (PD) is a complicated subject. Reasons for this complexity stem from the progressive nature of PD, its propensity to include neurobehavioral changes, and the variability in motor and nonmotor results that follow various types of surgical treatment. Studies of postsurgical behavioral changes also have not had the benefit of comparison with randomized, controlled, longitudinal follow-up of patients who have not had surgical treatment. Nevertheless, it appears that most modern-day lesioning and deep-brain stimulation (DBS) operations performed at experienced surgical centers are relatively safe from a cognitive perspective. Surgical interventions appear to involve minimal psychiatric morbidity in most instances. When neurocognitive declines are observed, they most often involve verbal fluency, regardless of the surgical technique and target. This decline in verbal fluency occurs more often after dominant hemisphere-sided operations and may persist for 1 year or more, independent of motor speech changes. Other selective changes have been observed in working memory, attention, and episodic memory, but only in a minority of studies and patients.
A small number of studies have reported severe cognitive and psychiatric complications, including both severe declines in memory and dementia. Although it is impossible to predict definitively which factors determine such declines, the following likely increase the risk for postoperative neurobehavioral morbidity: age (particularly > 70 years), bilateral or language-dominant hemisphere surgery (not necessarily ablative), lesion location in the anteromedial versus posterolateral aspect of the globus pallidus interna (GPi), pre-existing psychiatric disturbance, and preexisting dementia or marked frontal-subcortical syndrome. Lesion volume has not convincingly or consistently been identified as a correlate of cognitive outcome. Claims that DBS is neuropsychologically safer than ablative surgery are not adequately supported in the literature at this time. There is also no clear indication that certain targets (GPi versus subthalamic nucleus [STN]) are better from a cognitive standpoint. Future randomized, preferably blinded, studies are needed to compare interventions (target and treatment methods) directly in terms of their effects on neurobehavioral functioning.